Virtual patient simulation: Knowledge gain or knowledge loss?

Background: Virtual patients (VPs), high-fidelity simulators and standardized patients are powerful educational interventions leading to effective learning and supporting knowledge retention. Aim: This study explored the variations in retention with VP versus regular learning activities. Method: We conducted a randomized controlled study on early and delayed assessment results of 49 students using VP for learning and examination of haematology and cardiology topics in an Internal Medicine course, by means of a 0-10 scoring rubric. Results: The mean difference for early assessment with VP (study-control mean score) was 1.43 (95% confidence interval (CI) 0.96, 1.91; p < 0.001) for haematology and 1.34 (95% CI 0.93, 1.76; p < 0.001) for cardiology. In regular exams, the mean score difference was 2.21 (95% CI 1.3, 3.1; p < 0.001) and 1.52 (95% CI 0.76, 2.28; p < 0.001), respectively. With delayed assessments, the difference in mean score for Web-SP was 1.48 (95% CI 1.09, 1.86; p < 0.001), haematology and 1.16 (95% CI 0.74, 1.58; p < 0.001), cardiology; for regular exams the figures were 1.96 (95% CI 0.93, 2.98; p < 0.001) and 1.74 (95% CI 0.89, 2.58; p < 0.001). The effect size ranged from 0.5 to 0.8. Conclusion: Our results indicate better retention with VP than with traditional learning methods.